Patents
Literature
Patsnap Eureka AI that helps you search prior art, draft patents, and assess FTO risks, powered by patent and scientific literature data.

472results about How to "Avoid restrictions" patented technology

Method for predicting fatigue life of steel box girder by considering time-varying vehicle load and bridge model

The invention belongs to the field of bridge structure health monitoring, and discloses a steel box girder fatigue life prediction method considering a time-varying vehicle load and a bridge model, and the method comprises the following steps: carrying out the modeling of the time-varying fatigue vehicle load considering time-space distribution, including a random traffic flow feature model and a traffic flow long-term trend model; multi-scale model correction and calibration based on random vehicle-induced response, including modeling based on substructure-fine unit combination and bridge input and output relation reconstruction based on random vehicle-induced response probability characteristics; and loading a load on the calibrated bridge model to obtain a fatigue load effect, and dynamically predicting the fatigue life of the steel box girder based on the time-varying reliability. According to the method, fatigue accumulated damage in the service period can be considered, long-term dynamic evaluation of the fatigue performance of the steel box girder becomes possible based on the long-term monitoring calibration load model and the bridge loading model, and the method is easy to operate and has high engineering use value.
Owner:DALIAN UNIV OF TECH +2

Wind power equipment coating corrosion state monitoring method, device, equipment, storage medium and program product

The invention discloses a wind power equipment coating corrosion state monitoring method and device, equipment, a storage medium and a program product, and relates to the technical field of wind power generation, and the wind power equipment coating corrosion state monitoring method comprises the steps: obtaining equipment parameters of offshore wind power equipment and environment parameters collected by a sensor on the offshore wind power equipment; based on the equipment parameters and the environment parameters, establishing a digital twinborn model corresponding to the offshore wind power equipment; according to the digital twinborn model, the change condition of the coating state of the offshore wind power equipment is simulated, predicted coating state parameters are output, and the predicted coating state parameters at least comprise the coating corrosion state and the coating thickness of the offshore wind power equipment at the current time point. The technical problems that an existing wind power equipment coating corrosion state monitoring scheme is greatly limited by the environment, the workload is high, and the detection cost is high are solved.
Owner:SHENZHEN GUONENG CHENTAI TECH CO LTD

Medical image classification method based on prior knowledge enhanced mask and alignment modeling

The invention belongs to the technical field of computer vision, and particularly relates to a medical image classification method based on priori knowledge enhanced masks and alignment modeling. Comprising the following steps: acquiring an image text pair data set and preprocessing the image text pair data set to obtain a preprocessed image text pair data set; performing preliminary training on a visual encoder and a text encoder according to the preprocessed image text pair data set to obtain a preliminarily trained visual encoder and a preliminarily trained text encoder; training the visual encoder for the second time in combination with the text encoder to obtain a trained visual encoder; classifying the medical images by adopting the trained visual encoder; the model provided by the invention can be better generalized to various downstream medical image recognition tasks, and the accuracy of the downstream medical image recognition tasks is improved.
Owner:CHONGQING UNIV OF POSTS & TELECOMM

Virtual machine cross-operating system communication method based on microkernel Hypervisor

The invention discloses a micro-kernel Hypervisor-based virtual machine cross-operating system communication method, which comprises the following steps of: creating virtual machines which independently run in respective operating system environments by a micro-kernel Hypervisor, then constructing a shared memory between two virtual machines with communication requirements, respectively initializing virtual machine shared equipment by the two virtual machines, and storing the virtual machine shared equipment in the shared memory; according to the cross-operating-system communication method, the shared memory is mapped to a virtual address of the shared memory, when one virtual machine needs to transmit data to the opposite side, communication is initiated to write the data into the shared memory, and the virtual machine receiving the data reads the data from the shared memory, so that cross-operating-system communication of the virtual machines is realized. Based on the shared memory technology, the communication process of the shared memory and the virtual machines is managed by using the microkernel Hypervisor, and cross-operating system communication between the virtual machines is realized.
Owner:UNIV OF ELECTRONICS SCI & TECH OF CHINA +1

Attack trapping method and system based on network attack surface adaptive conversion

The invention discloses an attack trapping method and system based on network attack surface adaptive conversion, and the method comprises the steps: dynamically converting vulnerabilities and weaknesses in a system, so that an attacker is difficult to determine an entrance of a real system; threat intelligence and a network environment are monitored and analyzed in real time, and abnormal traffic and potential attack behaviors are identified; constructing a false target and monitoring attack behaviors, and obtaining behavior characteristics and attack means of an attacker; attack behavior data are analyzed by using artificial intelligence and a machine learning algorithm, potential threats are quickly identified and responded, a threat intelligence library and a defense strategy are updated, and the method better adapts to constantly changing attack means and threat forms. The method not only can trap attack behaviors, but also can collect and analyze behavior characteristics of attackers, and provides more accurate data support for network security decision and threat prediction.
Owner:BEIJING AEROSPACE WANYUAN TECH CO LTD

Reservoir intelligent scheduling method based on near-end strategy optimization algorithm

The invention discloses an intelligent reservoir scheduling method based on a near-end strategy optimization algorithm, and relates to the technical field of reservoir application. According to the method, a reservoir is modeled into an intelligent agent, the intelligent agent is composed of two neural networks of the same structure and interacts with a scheduling environment, network parameters are updated by adopting a near-end strategy optimization (PPO) algorithm, simulated reservoir inflow and forecast flow data are generated through an ARMA model according to historical actually-measured runoff data, and the intelligent agent is trained; the real-time scheduling performance of the intelligent agent is verified by using historical actually measured runoff data, behavior actions are expressed in a probability distribution form, the limitation of dimensionality disaster of a traditional algorithm is effectively avoided, and the Agent can more flexibly adjust a scheduling strategy by observing the water level and the incoming water condition in real time, so that the decision process is closer to an actual scheduling scene.
Owner:CHONGQING JIAOTONG UNIV +1

Optical proximity effect correction method of ion implantation layer and semiconductor device manufacturing method

The invention provides an optical proximity effect correction method of an ion implantation layer and a semiconductor device manufacturing method, and the method comprises the steps: firstly cutting off a convex corner in a corner-to-corner structure with a small interval in an original layout of the ion implantation layer, and then carrying out the OPC of the layout after the corner cutting, according to the method, the limitation of mask rule inspection on the corner-to-corner structures can be avoided, and the space for correcting the optical proximity effect is enlarged, so that the key size of a final layout obtained by OPC can meet a target value, the problem of process hot spots caused by easy distortion of a corner-to-corner structure pattern in an ion implantation layer in the prior art can be solved, and the quality of the ion implantation layer is improved. And the accuracy of OPC correction is improved, so that a more accurate final layout of the ion implantation layer is provided. According to the manufacturing method of the semiconductor device, the final layout of the required ion implantation layer is obtained by using the optical proximity effect correction method of the ion implantation layer, so that the yield and the performance of a product can be finally improved.
Owner:RONGXIN SEMICONDUCTOR (NINGBO) CO LTD

Method, device and equipment for preventing mistaken stepping of automobile accelerator pedal, medium and automobile

The invention relates to the technical field of vehicle control, and discloses a mistaken stepping prevention method, device and equipment for an automobile accelerator pedal, a medium and an automobile. The method comprises the steps that after it is detected that the opening degree of the accelerator pedal is changed, whether a barrier vehicle exists in front of a driving lane of the automobile or not is judged; if the obstacle vehicle exists, predicting whether the user executes lane changing overtaking according to the driving information of the vehicle; if not, judging whether the accelerator pedal is mistakenly stepped according to the relative driving state of the automobile and the obstacle vehicle and the opening information of the accelerator pedal; if it is judged that the accelerator pedal is not stepped by mistake, power output is conducted according to the accelerator opening degree corresponding to the accelerator pedal, and if it is judged that the accelerator pedal is stepped by mistake, power output is controlled. The driving intention of the driver can be predicted through multi-parameter comprehensive judgment, so that whether the accelerator pedal is mistakenly stepped or not is accurately judged based on the driving intention, the accelerator pedal is prevented from being limited when not mistakenly stepped, and meanwhile, the driving safety is timely guaranteed when the accelerator pedal is mistakenly stepped.
Owner:SUNGIANT AUTOMOTIVE ELECTRONICS CO LTD

Video hierarchical coding and decoding system and method based on region perception

The invention discloses a video hierarchical coding and decoding system and method based on area perception, and the method comprises the steps: an area perception stage: dividing an obtained monitoring video into video segments, processing video frames based on the video segments, and recognizing the area information of subsequent continuous frames through a reference frame processing result fed back by a server, the frame set of each video segment is spatially divided into a plurality of video sub-segments of different levels according to the application of the video; a coding stage: coding the video sub-segments into video sub-streams with different qualities by using different configuration parameters; and a decoding stage: decoding and restoring the video sub-streams into frames corresponding to the video sub-segments, and combining the frames into video frames. According to the invention, the time redundancy and the space redundancy of the video are fully considered, and the bandwidth of the transmitted video is greatly reduced; besides, the realized regional sensing mechanism overcomes the limitation of camera resources, not only meets the real-time processing requirement of video application, but also can be deployed on most cameras without GPU resources.
Owner:NANJING UNIV

Driving skill level classification method based on multi-dimensional data fusion

The invention discloses a driving skill level classification method based on multi-dimensional data fusion. The method comprises the following steps: (1) acquiring behaviors, physiological data and environmental data of a driver in real time through a driving simulator, eye movement tracking glasses and a physiological sensor; and (2) processing and analyzing the multi-dimensional data by using a CNN-LSTM classifier (CNN-LSTM-attention) so as to evaluate the skill level of the driver. According to the method, control skills and cognitive skills are comprehensively evaluated, key parameters including steering, braking, eye movement frequency, concentration duration and the like are captured in real time by using a driving simulator, eye movement tracking glasses and a physiological sensor, and a CNN-LSTM-attention classifier is constructed. According to the method, the complexity of the scene is quantified, the classifier is trained according to different scenes, the accuracy is improved, and meanwhile, the reliability of driving ability evaluation is enhanced by using multi-dimensional data.
Owner:TONGJI UNIV

Target detection method, device and equipment based on multi-sensor fusion and storage medium

The invention relates to the technical field of target detection, in particular to a target detection method, device and equipment based on multi-sensor fusion and a storage medium, and the method comprises the steps: tracking a target in real time through a millimeter wave radar, a photosensitive sensor and an infrared sensor, and judging the state of the target. According to the invention, the method can achieve the recognition of environment interference through the fusion of the three types of sensor data, has more intelligent environment adaptability, has excellent adaptability in different weathers and different scenes, can obtain more comprehensive target information through the mutual complementation of the three types of sensor data, and improves the detection accuracy and reliability.
Owner:SHENZHEN CHENG TECH CO LTD

Sewage treatment process optimization method and device based on deep learning

The invention discloses a sewage treatment process optimization method and device based on deep learning, and the method comprises the steps: carrying out the sample generation based on a generative adversarial network algorithm of hybrid differential privacy, and employing a reverse feature refinement strategy to adjust the data features in the sample generation process; performing feature extraction by adopting a full-connection neural network based on an adaptive wolf pack optimization algorithm; adopting a self-encoding neural network based on mask enhancement to selectively hide partial features of input data in an input layer by adopting a random mask to perform data compression and reconstruction so as to realize feature dimension reduction; and adopting a random forest algorithm based on model capacity self-adaptive pruning to perform classification prediction on the data after feature dimension reduction to obtain a classification result of the sewage treatment degree. According to the method, the sewage treatment process is integrally optimized and improved, the sewage treatment process can be monitored and dynamically optimized in real time after improvement, the sewage treatment efficiency is improved, the sewage treatment effect is improved, and it is ensured that the sewage treatment process meets the environmental protection standard.
Owner:YUNNAN AGRICULTURAL UNIVERSITY

Non-contact physiological and psychological detection system based on multiple modes

The invention discloses a non-contact physiological and psychological detection system based on multiple modes, relates to the cross technical field of artificial intelligence and biomedical engineering, and can collect multi-mode data such as facial images and sound signals of a user at the same time by integrating multiple sensors such as a camera and a microphone. The fusion of the multi-source data enables the system to more comprehensively capture the physiological and psychological state changes of the user, thereby improving the accuracy and reliability of evaluation. The system collects data in a non-contact mode, and discomfort and limitation possibly brought by traditional physiological monitoring equipment are avoided. According to the system, various non-contact detection technologies such as computer vision, voice analysis and natural language processing are comprehensively applied, and the physiological and psychological states of the user can be evaluated from multiple dimensions. The non-contact detection technology reduces the interference to the user, so that the detection process is more natural and efficient.
Owner:黎卫兵

Internet of Things HTTP fuzzy testing method based on large language model

The invention discloses an Internet of Things HTTP fuzzing method based on a large language model, and relates to the technical field of Internet of Things HTTP fuzzing. The method comprises the following steps: data acquisition: adopting a data acquisition strategy to obtain data for the HTTP fuzzy test of the Internet of Things; data preprocessing: preprocessing the collected data, and providing high-quality input data for subsequent fuzzy testing; hTTP structure sensing guided by LLM: using a prompt adjustment method to label fuzzy variables in request data by using the LLM, understanding a protocol data structure, and guiding a variation process; lLM-guided seed template expansion: expanding a seed bank by enriching data packets and expanding field values; lLM-guided seed variation and seed scheduling: optimizing test results by intelligently selecting and managing input seeds. According to the method, the efficiency and accuracy of fuzzy testing can be improved, and the vulnerability detection capability is enhanced.
Owner:SHIJIAZHUANG TIEDAO UNIV

Flow field numerical simulation method for autonomous swimming of underwater bionic fish

The invention discloses a flow field numerical simulation method for autonomous swimming of underwater bionic fish, and particularly relates to the field of bionic fish swimming flow field numerical simulation, which comprises the step of describing the shape of a fish body by establishing a contour line equation of a horizontal section. The global ground coordinate system and the fish body coordinate system are associated through the gravity center position and the pitch angle of the fish body, and a time sequence fish body motion data set is established. And generating a time sequence flow field data set by combining unsteady flow field data acquired by the distributed underwater sensor array. And based on the flow field control equation and the dynamic motion characteristics of the fish body, constructing a static interactive coupling model, extracting key dynamic characteristic quantities, and establishing a flow field prediction model. And inputting interactive coupling data to the static model to calculate a global coupling solution for error analysis, and generating a dynamic interactive coupling model. And according to the flow field prediction result and the dynamic interactive coupling model, correcting the motion characteristic data of the fish body, optimizing the motion trail of the bionic fish, and realizing closed-loop optimization of the flow field and the fish body motion.
Owner:NANJING UNIV OF AERONAUTICS & ASTRONAUTICS +1

Heterogeneous graph-based man-post matching method and device, equipment and medium

The invention discloses a person and post matching method and device based on a heterogeneous graph, equipment and a medium. Comprising the steps of obtaining position information, resume information and historical matching behavior data between the position information and the resume information; taking the position information as a position node, taking the resume information as a resume node, and establishing a first edge relationship between the position node and the resume node based on historical matching behavior data; based on the similarity of each piece of position information, the similarity of each piece of resume information and historical matching behavior data, establishing a second edge relationship between position nodes, a third edge relationship between resume nodes and a fourth edge relationship between the position nodes and the resume nodes; constructing a global heterogeneous graph based on each position node, each resume node and the edge relationship thereof; training a GLAN model based on the global heterogeneous graph; and converting the position information to be matched into heterogeneous graph features based on the GLAN model, and screening matched target resume information for the position information to be matched based on the heterogeneous graph features.
Owner:QIAN JIN NETWORK INFORMATION TECH SHANGHAI LTD

Indirect prediction method of milling force considering tool wear based on power signal

The present invention provides an indirect prediction method for milling force based on power signal considering tool wear. By establishing a theoretical power consumption model and considering the influence of tool wear on cutting force, the influence relationship from tool wear to cutting force and then to machine tool power is established from the mechanism level, and a variable working condition milling force prediction model based on machine tool power consumption is established, thereby realizing accurate prediction of tool wear and cutting force under any working condition. The method provided by the present invention overcomes the shortcomings of other indirect monitoring signals in terms of versatility and extensibility. At the same time, the present invention is relatively simple to install and maintain, and has higher feasibility in actual production line deployment.
Owner:NORTHWESTERN POLYTECHNICAL UNIV

Meteorological data analysis method, device and equipment based on large model agent and medium

The invention discloses a meteorological data analysis method, device and equipment based on a large model agent and a medium, relates to the field of meteorological data intelligent management, is applied to a preset meteorological data analysis system, and comprises the following steps: when a main agent receives a to-be-processed meteorological message, determining and calling a target agent from a plurality of extreme weather scene agents; the main agent and the plurality of extreme weather scene agents are agents constructed by using a target large language model in advance; the target agent matches a first cue word template of the target agent based on a received main agent instruction so as to determine and call a target meteorological data acquisition tool; the target meteorological data acquisition tool matches a second cue word template of the target meteorological data acquisition tool based on a received target agent instruction so as to acquire meteorological data based on the determined target interface parameters; and the target agent determines a target answer based on the data acquisition result, so that the main agent replies a message in a dialogue form. And the limitation of the existing meteorological data application can be overcome.
Owner:SHANDONG LANGCHAO YUNTOU INFORMATION TECH CO LTD

Data communication method, device, equipment and medium

The invention discloses a data communication method, a data communication device, data communication equipment and a medium, relates to the technical field of computers, and is applied to a complex programmable logic device which communicates with a south bridge chip in a preset type server and an external target substrate management controller chip through a simplified pin bus. Comprising the steps that target data sent by a south bridge chip is transmitted to a target substrate management controller chip, and after a TAR signal is monitored, a first synchronous communication signal fed back by the target substrate management controller chip is acquired; converting the first synchronous communication signal into a second synchronous communication signal; and transmitting the second synchronous communication signal to a south bridge chip, so that the south bridge chip communicates with the target substrate management controller chip based on the simplified pin bus after executing the preset short-time waiting operation. According to the application, the communication between the preset type server and the target substrate management controller chip based on the simplified pin bus is realized by connecting the complex programmable logic device in series.
Owner:INSPUR (SHANDONG) COMPUTER TECH CO LTD

Application of antioxidant lipidosome in preparation of medicine for preventing and / or treating inner ear injury caused by active oxygen accumulation

The invention relates to application of an antioxidant liposome in preparation of a medicine for preventing and / or treating inner ear injury caused by active oxygen accumulation, and belongs to the field of medical technologies and biological medical materials. The antioxidant lipidosome is rutin-entrapped lipidosome, and the inner ear injury is preferably cis-platinum induced inner ear injury. The antioxidant liposome disclosed by the invention is used for inhibiting active oxygen accumulation, mitochondrial injury and cell apoptosis in cells; the compound is used for inhibiting auditory hair cell mitochondrial injury in cochlea. The antioxidant liposome disclosed by the invention can be used for effectively packaging medicines and improving the bioavailability and the stability of the medicines; the limitation of a tympanic membrane barrier and other biological barriers can be overcome, so that the medicine can effectively reach the inner ear part; precise targeted delivery of the medicine can be achieved, and it is ensured that the medicine is intensively released at the needed position.
Owner:XIEHE HOSPITAL ATTACHED TO TONGJI MEDICAL COLLEGE HUAZHONG SCI & TECH UNIV

Network intrusion detection method based on depth time sequence clustering

PendingCN119135404Aavoid restrictionsAvoid the problem of needing to determine the number of kBiological modelsSecuring communicationData setSupervised learning
The invention provides a network intrusion detection method based on depth time sequence clustering. The method comprises the following steps: performing dimensionality reduction and reconstruction on high-dimensional data by using a convolutional automatic encoder CAE; the self-supervised learning module performs feature transformation on the data, the training model classifies and identifies the transformation type of a sample as an auxiliary task, and learns generalized feature representation; global and local memory modules learn universal and specific features of data; the adaptive fusion module fuses the features to obtain a new feature representation for reconstruction; calculating the most suitable clustering feature of the positive sample data set; and identifying data different from the positive sample data set in the low-dimensional space by using a BIRCH clustering algorithm to realize anomaly detection. The self-supervised learning module is introduced to avoid the problem of data source limitation, the global and local memory modules are utilized to avoid the problems of poor model generalization ability and feature expression limitation, the BIRCH clustering is utilized to avoid the problem that the quantity of k needs to be determined in the traditional K-MEANS algorithm, and the method has higher efficiency.
Owner:YANCHENG INST OF TECH +1

Navigation point correction method and system and photovoltaic cleaning device

The invention relates to the field of photovoltaic cleaning, and discloses a navigation point correction method and system and a photovoltaic cleaning device.The method comprises the steps that a multi-line laser radar arranged on the photovoltaic cleaning device is used for collecting point cloud data of a target photovoltaic panel; constructing an annular point cloud picture by taking the multi-line laser radar as a center, and partitioning the annular point cloud picture; obtaining representative points of each partition; the representative points of the adjacent partitions are connected, and edge point clouds of the target photovoltaic panel are screened out; performing linear fitting on the screened edge point cloud; and calculating the position of the latest target navigation point in a coordinate system with the multi-line laser radar as the center in real time by using the edge straight line obtained by fitting. Through the point cloud partitioning technology, the expression equation of the edge of the photovoltaic panel is detected in real time, no matter how the inclination angle of the photovoltaic panel changes, the cleaning device can keep a fixed cleaning distance from the photovoltaic panel all the time, and therefore the cleaning efficiency and safety are improved.
Owner:LEAPTING TECH CO LTD

Tower type high-efficiency granulation device for compound fertilizer production

The tower-type compound fertilizer production efficient granulation device comprises a tower body, separation seats are symmetrically installed on the inner wall of the tower body, a plurality of separation pipes are installed at the position, located between the two separation seats, in the tower body in the circumferential direction at equal angles, and piston boxes are symmetrically installed on the end face of one side of the tower body; a rotational flow box is installed in the middle of the side end face of the piston box, a movable piston is slidably installed in the piston box, and a rotating wheel is rotatably installed in the rotational flow box. Driving force comes from slurry heat and heat in air, limitation of external energy to the device can be effectively avoided, the device can operate more efficiently and stably, and the device is more stable in operation. The device is more energy-saving and environment-friendly, the processing cost of the compound fertilizer is reduced, the heat carried by the slurry is actively extracted through gas-liquid conversion, the influence of external solar terms and regions on the device is effectively avoided while the heat exchange efficiency is improved, and the device can perform granulation processing more freely and stably.
Owner:HEBEI DUODALIANG ECOLOGICAL FERTILIZER CO LTD

Display module and terminal equipment

The utility model provides a display module and terminal equipment, the display module comprises a display panel and a driver, the display panel comprises a display part, a bending part and a turnover part, the bending part is connected between the display part and the turnover part, the turnover part is located on one side of the display part in the thickness direction and is arranged opposite to the display part, and the driver is arranged on the turnover part. The folding part comprises a first part and a second part, the first part is connected between the bending part and the second part, the first part comprises a first surface in the thickness direction, the first surface deviates from the display part, the second part comprises a second surface in the thickness direction, the second surface deviates from the display part, and the driver is installed on the second surface of the second part; the second surface is located on the side, facing the display part, of the first surface. The display module is locally thinned, and the thickness bottleneck problem of the display module is solved.
Owner:HONOR DEVICE CO LTD

Optical element surface defect data set manufacturing method based on deep learning

PendingCN119580035AAvoid the contradiction of large demand but small actual quantityavoid restrictionsImage analysisCharacter and pattern recognitionPattern recognitionData set
The invention discloses an optical element surface defect data set manufacturing method based on deep learning. The method comprises the following steps: acquiring an optical element surface defect image; inputting the obtained defect image into a pre-trained defect detection model to carry out defect automatic detection; based on the defect detection result, drawing the outer contour of each defect by using a contour retrieval method, counting the corresponding defect type and defect total number, and then generating a contour retrieval result; carrying out rapid manual correction on the obtained contour retrieval result; and generating a label of the optical element surface defect image. The defect detection model is a semantic segmentation network, can not only judge whether defects exist or not and perform defect classification, but also can realize defect positioning and extraction at a pixel level. According to the method, the probability of false detection and missing detection under a complex background when the defect pixel-level label is manufactured is reduced, and the manufacturing efficiency and the manufacturing precision of the surface defect data set of the optical element are greatly improved.
Owner:INST OF OPTICS & ELECTRONICS CHINESE ACAD OF SCI

Professional domain database retrieval method based on big language model knowledge enhancement, electronic equipment and medium

The invention discloses a professional domain database retrieval method based on big language model knowledge enhancement, electronic equipment and a medium, and the method comprises the following steps: obtaining keywords and descriptions thereof corresponding to each type of data in a target domain database, and matching the keywords with data links and contents to obtain an SQL (Structured Query Language) table; defining a first corpus template, and serially connecting keywords and description thereof to obtain a data prompt text; defining a second corpus template, and connecting the query texts in series to obtain a query prompt text; respectively inputting the data and the query prompt text into a large language model, and respectively generating a response text and a retrieval text; inputting the query text, the retrieval text and the response text into a natural language processing model to generate query text features, retrieval text features and response text features; performing weighted fusion on the query and retrieval text features to obtain fused text features; and calculating the similarity between the first K response text features and the response text features, and exporting data corresponding to the first K response text features with the highest similarity from the SQL table as a retrieval result.
Owner:ZHEJIANG LAB

Flood reservoir flow prediction method and device and related equipment

The invention provides a flood reservoir flow prediction method and device and related equipment, and relates to the technical field of disaster prevention and mitigation, and the method comprises the steps: obtaining a plurality of training data pairs corresponding to a target water conservancy facility, the plurality of training data pairs respectively corresponding to a plurality of floods experienced by the target water conservancy facility in a historical period, the training data pair comprises rainfall data indicating the rainfall of the corresponding flood and storage data indicating the storage flow of the target water conservancy facility in the corresponding flood; the in-storage data is data obtained through flood regulation calculation back calculation under the condition of water balance based on the out-storage data of the corresponding flood and the storage capacity data of the corresponding flood; performing data analysis on the plurality of training data pairs to obtain a target mapping relation; and according to the target mapping relationship and the rainfall of the current flood, predicting reservoir flow information of the current flood in the target water conservancy facility. According to the invention, more accurate storage flow information can be predicted.
Owner:ZHEJIANG INST OF HYDRAULICS & ESTUARY

Method for non-contact measurement of gap between open-type TBM shield tail and surrounding rock

The invention belongs to the field of TBM shield tail gap measurement, and particularly relates to a method for non-contact measurement of a gap between an open TBM shield tail and a surrounding rock, and the method combines Blender software to establish a virtual scene comprising the surrounding rock, a TBM shield, a reflective sticker, a binocular camera and an LED spotlight. The textures of the TBM shield, the surrounding rock and the reflective sticker in the virtual scene can be set according to a real image, a large number of binocular camera images can be rendered through changes of the textures and the surrounding rock surface, meanwhile, a gap between the tail of the TBM shield and the surrounding rock under a corresponding binocular graph is obtained through a written blender script, and a data set is generated. The data set is used for building a neural network model on a server, training is carried out, and a training weight is obtained; and then deploying the model and the weight into a measuring device, namely shooting an image of the binocular camera through the measuring device, and measuring the gap between the tail part of the shield and the surrounding rock.
Owner:CHINA RAILWAY SHISIJU GROUP CORP +1

Planting device with infiltrating irrigation function

The utility model discloses a planting device with an infiltrating irrigation function, which comprises a planter, a water pump, a water pump and a water pump, a cavity for planting flowers and trees is arranged in the planter, and a mounting hole is arranged on the side wall or the bottom wall of the planter; the water supply pipe group is fixedly connected with the mounting hole; the water seepage assembly comprises a connecting base and a plurality of membrane wires, the ends of the membrane wires are fixed to the connecting base in a bunch mode, the membrane wires are dispersed in the containing cavity and close to the root system of the plant, and water flow in the water supply pipe set permeates outwards through the membrane wires and wets soil. The problem that urban greening or road green belts waste water sources is solved, the cultivation limitation of the severe environment on green plants is reduced, greening and environment recovery of places such as stone mountains, rugged areas and arid areas where watering is inconvenient are facilitated, and greening of family courtyards and balconies can be facilitated.
Owner:SHANXI ZEYUAN WATER SAVING IRRIGATION ENGINEERING TECHNOLOGY CO LTD

Barb type fastener

The utility model discloses a barb type fastener, and relates to the technical field of hardware fittings. The barb type fastener comprises a metal composite plate body, an inner side mounting part and a positioning structure, a first anti-falling groove is formed in the outer wall of the metal composite plate body, a second anti-falling groove is formed in the inner wall of the metal composite plate body, the inner side mounting part is arranged on the inner wall of the metal composite plate body, and the positioning structure is arranged on the inner side mounting part. Two groups of connecting sleeves are fixedly mounted on the outer wall of the inner side mounting part, one end of each connecting sleeve is fixedly connected with an outer side mounting part, and the positioning structures are arranged on the inner side mounting part and the outer side mounting part. According to the barb type fastener, follow-up installation can be facilitated only by prefabricating the first anti-disengagement groove and the second anti-disengagement groove in the engraving operation stage of the A-level non-combustible aluminum composite board, the installation process is simple and convenient, a large amount of manual operation does not need to be consumed, the A-level non-combustible metal composite board can be conveniently and rapidly produced, and the production efficiency is improved. And the practicability of the fastener is effectively enhanced.
Owner:LEANGU NEW MATERIALS (MAANSHAN) CO LTD